]> Raphaël G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/yahoo.py
Merge tag 'upstream/2016.12.01'
[youtubedl] / youtube_dl / extractor / yahoo.py
index f8e7041a08d042ac44c13338439b5568bf4caac6..4951414e91ffbc34dc83629403c8b64ffb5e5682 100644 (file)
@@ -15,16 +15,23 @@ from ..utils import (
     unescapeHTML,
     ExtractorError,
     int_or_none,
+    mimetype2ext,
+    determine_ext,
 )
 
+from .brightcove import (
+    BrightcoveLegacyIE,
+    BrightcoveNewIE,
+)
+from .nbc import NBCSportsVPlayerIE
+
 
 class YahooIE(InfoExtractor):
     IE_DESC = 'Yahoo screen and movies'
-    _VALID_URL = r'(?P<url>(?P<host>https?://(?:[a-zA-Z]{2}\.)?[\da-zA-Z_-]+\.yahoo\.com)/(?:[^/]+/)*(?P<display_id>.+?)-(?P<id>[0-9]+)(?:-[a-z]+)?\.html)'
+    _VALID_URL = r'(?P<url>(?P<host>https?://(?:[a-zA-Z]{2}\.)?[\da-zA-Z_-]+\.yahoo\.com)/(?:[^/]+/)*(?P<display_id>.+)?-(?P<id>[0-9]+)(?:-[a-z]+)?(?:\.html)?)'
     _TESTS = [
         {
             'url': 'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html',
-            'md5': '4962b075c08be8690a922ee026d05e69',
             'info_dict': {
                 'id': '2d25e626-2378-391f-ada0-ddaf1417e588',
                 'ext': 'mp4',
@@ -35,7 +42,7 @@ class YahooIE(InfoExtractor):
         },
         {
             'url': 'http://screen.yahoo.com/wired/codefellas-s1-ep12-cougar-lies-103000935.html',
-            'md5': 'd6e6fc6e1313c608f316ddad7b82b306',
+            'md5': '251af144a19ebc4a033e8ba91ac726bb',
             'info_dict': {
                 'id': 'd1dedf8c-d58c-38c3-8963-e899929ae0a9',
                 'ext': 'mp4',
@@ -46,7 +53,7 @@ class YahooIE(InfoExtractor):
         },
         {
             'url': 'https://screen.yahoo.com/community/community-sizzle-reel-203225340.html?format=embed',
-            'md5': '60e8ac193d8fb71997caa8fce54c6460',
+            'md5': '7993e572fac98e044588d0b5260f4352',
             'info_dict': {
                 'id': '4fe78544-8d48-39d8-97cd-13f205d9fcdb',
                 'ext': 'mp4',
@@ -56,22 +63,22 @@ class YahooIE(InfoExtractor):
             }
         },
         {
-            'url': 'https://tw.screen.yahoo.com/election-2014-askmayor/敢問市長-黃秀霜批賴清德-非常高傲-033009720.html',
-            'md5': '3a09cf59349cfaddae1797acc3c087fc',
+            'url': 'https://tw.news.yahoo.com/%E6%95%A2%E5%95%8F%E5%B8%82%E9%95%B7%20%E9%BB%83%E7%A7%80%E9%9C%9C%E6%89%B9%E8%B3%B4%E6%B8%85%E5%BE%B7%20%E9%9D%9E%E5%B8%B8%E9%AB%98%E5%82%B2-034024051.html',
+            'md5': '45c024bad51e63e9b6f6fad7a43a8c23',
             'info_dict': {
                 'id': 'cac903b3-fcf4-3c14-b632-643ab541712f',
                 'ext': 'mp4',
                 'title': '敢問市長/黃秀霜批賴清德「非常高傲」',
                 'description': '直言台南沒捷運 交通居五都之末',
                 'duration': 396,
-            }
+            },
         },
         {
             'url': 'https://uk.screen.yahoo.com/editor-picks/cute-raccoon-freed-drain-using-091756545.html',
-            'md5': '0b51660361f0e27c9789e7037ef76f4b',
+            'md5': '71298482f7c64cbb7fa064e4553ff1c1',
             'info_dict': {
                 'id': 'b3affa53-2e14-3590-852b-0e0db6cd1a58',
-                'ext': 'mp4',
+                'ext': 'webm',
                 'title': 'Cute Raccoon Freed From Drain\u00a0Using Angle Grinder',
                 'description': 'md5:f66c890e1490f4910a9953c941dee944',
                 'duration': 97,
@@ -86,20 +93,35 @@ class YahooIE(InfoExtractor):
                 'title': 'Program that makes hockey more affordable not offered in Manitoba',
                 'description': 'md5:c54a609f4c078d92b74ffb9bf1f496f4',
                 'duration': 121,
-            }
+            },
+            'skip': 'Video gone',
         }, {
             'url': 'https://ca.finance.yahoo.com/news/hackers-sony-more-trouble-well-154609075.html',
-            'md5': '226a895aae7e21b0129e2a2006fe9690',
             'info_dict': {
-                'id': 'e624c4bc-3389-34de-9dfc-025f74943409',
-                'ext': 'mp4',
-                'title': '\'The Interview\' TV Spot: War',
-                'description': 'The Interview',
-                'duration': 30,
-            }
+                'id': '154609075',
+            },
+            'playlist': [{
+                'md5': '000887d0dc609bc3a47c974151a40fb8',
+                'info_dict': {
+                    'id': 'e624c4bc-3389-34de-9dfc-025f74943409',
+                    'ext': 'mp4',
+                    'title': '\'The Interview\' TV Spot: War',
+                    'description': 'The Interview',
+                    'duration': 30,
+                },
+            }, {
+                'md5': '81bc74faf10750fe36e4542f9a184c66',
+                'info_dict': {
+                    'id': '1fc8ada0-718e-3abe-a450-bf31f246d1a9',
+                    'ext': 'mp4',
+                    'title': '\'The Interview\' TV Spot: Guys',
+                    'description': 'The Interview',
+                    'duration': 30,
+                },
+            }],
         }, {
             'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
-            'md5': '67010fdf3a08d290e060a4dd96baa07b',
+            'md5': '88e209b417f173d86186bef6e4d1f160',
             'info_dict': {
                 'id': 'f885cf7f-43d4-3450-9fac-46ac30ece521',
                 'ext': 'mp4',
@@ -116,10 +138,11 @@ class YahooIE(InfoExtractor):
                 'title': 'Connect the Dots: Dark Side of Virgo',
                 'description': 'md5:1428185051cfd1949807ad4ff6d3686a',
                 'duration': 201,
-            }
+            },
+            'skip': 'Domain name in.lifestyle.yahoo.com gone',
         }, {
             'url': 'https://www.yahoo.com/movies/v/true-story-trailer-173000497.html',
-            'md5': '989396ae73d20c6f057746fb226aa215',
+            'md5': '2a9752f74cb898af5d1083ea9f661b58',
             'info_dict': {
                 'id': '071c4013-ce30-3a93-a5b2-e0413cd4a9d1',
                 'ext': 'mp4',
@@ -130,40 +153,153 @@ class YahooIE(InfoExtractor):
         }, {
             'url': 'https://gma.yahoo.com/pizza-delivery-man-surprised-huge-tip-college-kids-195200785.html',
             'only_matching': True,
-        }
+        }, {
+            'note': 'NBC Sports embeds',
+            'url': 'http://sports.yahoo.com/blogs/ncaab-the-dagger/tyler-kalinoski-s-buzzer-beater-caps-davidson-s-comeback-win-185609842.html?guid=nbc_cbk_davidsonbuzzerbeater_150313',
+            'info_dict': {
+                'id': '9CsDKds0kvHI',
+                'ext': 'flv',
+                'description': 'md5:df390f70a9ba7c95ff1daace988f0d8d',
+                'title': 'Tyler Kalinoski hits buzzer-beater to lift Davidson',
+                'upload_date': '20150313',
+                'uploader': 'NBCU-SPORTS',
+                'timestamp': 1426270238,
+            }
+        }, {
+            'url': 'https://tw.news.yahoo.com/-100120367.html',
+            'only_matching': True,
+        }, {
+            # Query result is embedded in webpage, but explicit request to video API fails with geo restriction
+            'url': 'https://screen.yahoo.com/community/communitary-community-episode-1-ladders-154501237.html',
+            'md5': '4fbafb9c9b6f07aa8f870629f6671b35',
+            'info_dict': {
+                'id': '1f32853c-a271-3eef-8cb6-f6d6872cb504',
+                'ext': 'mp4',
+                'title': 'Communitary - Community Episode 1: Ladders',
+                'description': 'md5:8fc39608213295748e1e289807838c97',
+                'duration': 1646,
+            },
+        }, {
+            # it uses an alias to get the video_id
+            'url': 'https://www.yahoo.com/movies/the-stars-of-daddys-home-have-very-different-212843197.html',
+            'info_dict': {
+                'id': '40eda9c8-8e5f-3552-8745-830f67d0c737',
+                'ext': 'mp4',
+                'title': 'Will Ferrell & Mark Wahlberg Are Pro-Spanking',
+                'description': 'While they play feuding fathers in \'Daddy\'s Home,\' star Will Ferrell & Mark Wahlberg share their true feelings on parenthood.',
+            },
+        },
+        {
+            # config['models']['applet_model']['data']['sapi'] has no query
+            'url': 'https://www.yahoo.com/music/livenation/event/galactic-2016',
+            'md5': 'dac0c72d502bc5facda80c9e6d5c98db',
+            'info_dict': {
+                'id': 'a6015640-e9e5-3efb-bb60-05589a183919',
+                'ext': 'mp4',
+                'description': 'Galactic',
+                'title': 'Dolla Diva (feat. Maggie Koerner)',
+            },
+            'skip': 'redirect to https://www.yahoo.com/music',
+        },
+        {
+            # yahoo://article/
+            'url': 'https://www.yahoo.com/movies/video/true-story-trailer-173000497.html',
+            'info_dict': {
+                'id': '071c4013-ce30-3a93-a5b2-e0413cd4a9d1',
+                'ext': 'mp4',
+                'title': "'True Story' Trailer",
+                'description': 'True Story',
+            },
+            'params': {
+                'skip_download': True,
+            },
+        },
+        {
+            # ytwnews://cavideo/
+            'url': 'https://tw.video.yahoo.com/movie-tw/單車天使-中文版預-092316541.html',
+            'info_dict': {
+                'id': 'ba133ff2-0793-3510-b636-59dfe9ff6cff',
+                'ext': 'mp4',
+                'title': '單車天使 - 中文版預',
+                'description': '中文版預',
+            },
+            'params': {
+                'skip_download': True,
+            },
+        },
     ]
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
-        display_id = mobj.group('display_id')
+        display_id = mobj.group('display_id') or self._match_id(url)
         page_id = mobj.group('id')
         url = mobj.group('url')
         host = mobj.group('host')
-        webpage = self._download_webpage(url, display_id)
+        webpage, urlh = self._download_webpage_handle(url, display_id)
+        if 'err=404' in urlh.geturl():
+            raise ExtractorError('Video gone', expected=True)
 
         # Look for iframed media first
-        iframe_m = re.search(r'<iframe[^>]+src="(/video/.+?-\d+\.html\?format=embed.*?)"', webpage)
-        if iframe_m:
-            iframepage = self._download_webpage(
-                host + iframe_m.group(1), display_id, 'Downloading iframe webpage')
-            items_json = self._search_regex(
-                r'mediaItems: (\[.+?\])$', iframepage, 'items', flags=re.MULTILINE, default=None)
-            if items_json:
-                items = json.loads(items_json)
-                video_id = items[0]['id']
-                return self._get_info(video_id, display_id, webpage)
+        entries = []
+        iframe_urls = re.findall(r'<iframe[^>]+src="(/video/.+?-\d+\.html\?format=embed.*?)"', webpage)
+        for idx, iframe_url in enumerate(iframe_urls):
+            entries.append(self.url_result(host + iframe_url, 'Yahoo'))
+        if entries:
+            return self.playlist_result(entries, page_id)
+
+        # Look for NBCSports iframes
+        nbc_sports_url = NBCSportsVPlayerIE._extract_url(webpage)
+        if nbc_sports_url:
+            return self.url_result(nbc_sports_url, NBCSportsVPlayerIE.ie_key())
+
+        # Look for Brightcove Legacy Studio embeds
+        bc_url = BrightcoveLegacyIE._extract_brightcove_url(webpage)
+        if bc_url:
+            return self.url_result(bc_url, BrightcoveLegacyIE.ie_key())
+
+        # Look for Brightcove New Studio embeds
+        bc_url = BrightcoveNewIE._extract_url(webpage)
+        if bc_url:
+            return self.url_result(bc_url, BrightcoveNewIE.ie_key())
+
+        # Query result is often embedded in webpage as JSON. Sometimes explicit requests
+        # to video API results in a failure with geo restriction reason therefore using
+        # embedded query result when present sounds reasonable.
+        config_json = self._search_regex(
+            r'window\.Af\.bootstrap\[[^\]]+\]\s*=\s*({.*?"applet_type"\s*:\s*"td-applet-videoplayer".*?});(?:</script>|$)',
+            webpage, 'videoplayer applet', default=None)
+        if config_json:
+            config = self._parse_json(config_json, display_id, fatal=False)
+            if config:
+                sapi = config.get('models', {}).get('applet_model', {}).get('data', {}).get('sapi')
+                if sapi and 'query' in sapi:
+                    info = self._extract_info(display_id, sapi, webpage)
+                    self._sort_formats(info['formats'])
+                    return info
 
         items_json = self._search_regex(
             r'mediaItems: ({.*?})$', webpage, 'items', flags=re.MULTILINE,
             default=None)
         if items_json is None:
-            CONTENT_ID_REGEXES = [
-                r'YUI\.namespace\("Media"\)\.CONTENT_ID\s*=\s*"([^"]+)"',
-                r'root\.App\.Cache\.context\.videoCache\.curVideo = \{"([^"]+)"',
-                r'"first_videoid"\s*:\s*"([^"]+)"',
-                r'%s[^}]*"ccm_id"\s*:\s*"([^"]+)"' % re.escape(page_id),
-            ]
-            video_id = self._search_regex(CONTENT_ID_REGEXES, webpage, 'content ID')
+            alias = self._search_regex(
+                r'"aliases":{"video":"(.*?)"', webpage, 'alias', default=None)
+            if alias is not None:
+                alias_info = self._download_json(
+                    'https://www.yahoo.com/_td/api/resource/VideoService.videos;video_aliases=["%s"]' % alias,
+                    display_id, 'Downloading alias info')
+                video_id = alias_info[0]['id']
+            else:
+                CONTENT_ID_REGEXES = [
+                    r'YUI\.namespace\("Media"\)\.CONTENT_ID\s*=\s*"([^"]+)"',
+                    r'root\.App\.Cache\.context\.videoCache\.curVideo = \{"([^"]+)"',
+                    r'"first_videoid"\s*:\s*"([^"]+)"',
+                    r'%s[^}]*"ccm_id"\s*:\s*"([^"]+)"' % re.escape(page_id),
+                    r'<article[^>]data-uuid=["\']([^"\']+)',
+                    r'<meta[^<>]+yahoo://article/view\?.*\buuid=([^&"\']+)',
+                    r'<meta[^<>]+["\']ytwnews://cavideo/(?:[^/]+/)+([\da-fA-F-]+)[&"\']',
+                ]
+                video_id = self._search_regex(
+                    CONTENT_ID_REGEXES, webpage, 'content ID')
         else:
             items = json.loads(items_json)
             info = items['mediaItems']['query']['results']['mediaObj'][0]
@@ -172,22 +308,10 @@ class YahooIE(InfoExtractor):
             video_id = info['id']
         return self._get_info(video_id, display_id, webpage)
 
-    def _get_info(self, video_id, display_id, webpage):
-        region = self._search_regex(
-            r'\\?"region\\?"\s*:\s*\\?"([^"]+?)\\?"',
-            webpage, 'region', fatal=False, default='US')
-        data = compat_urllib_parse.urlencode({
-            'protocol': 'http',
-            'region': region,
-        })
-        query_url = (
-            'https://video.media.yql.yahoo.com/v1/video/sapi/streams/'
-            '{id}?{data}'.format(id=video_id, data=data))
-        query_result = self._download_json(
-            query_url, display_id, 'Downloading video info')
-
-        info = query_result['query']['results']['mediaObj'][0]
+    def _extract_info(self, display_id, query, webpage):
+        info = query['query']['results']['mediaObj'][0]
         meta = info.get('meta')
+        video_id = info.get('id')
 
         if not meta:
             msg = info['status'].get('msg')
@@ -198,26 +322,51 @@ class YahooIE(InfoExtractor):
 
         formats = []
         for s in info['streams']:
+            tbr = int_or_none(s.get('bitrate'))
             format_info = {
                 'width': int_or_none(s.get('width')),
                 'height': int_or_none(s.get('height')),
-                'tbr': int_or_none(s.get('bitrate')),
+                'tbr': tbr,
             }
 
             host = s['host']
             path = s['path']
             if host.startswith('rtmp'):
+                fmt = 'rtmp'
                 format_info.update({
                     'url': host,
                     'play_path': path,
                     'ext': 'flv',
                 })
             else:
+                if s.get('format') == 'm3u8_playlist':
+                    fmt = 'hls'
+                    format_info.update({
+                        'protocol': 'm3u8_native',
+                        'ext': 'mp4',
+                    })
+                else:
+                    fmt = format_info['ext'] = determine_ext(path)
                 format_url = compat_urlparse.urljoin(host, path)
                 format_info['url'] = format_url
+            format_info['format_id'] = fmt + ('-%d' % tbr if tbr else '')
             formats.append(format_info)
 
-        self._sort_formats(formats)
+        closed_captions = self._html_search_regex(
+            r'"closedcaptions":(\[[^\]]+\])', webpage, 'closed captions',
+            default='[]')
+
+        cc_json = self._parse_json(closed_captions, video_id, fatal=False)
+        subtitles = {}
+        if cc_json:
+            for closed_caption in cc_json:
+                lang = closed_caption['lang']
+                if lang not in subtitles:
+                    subtitles[lang] = []
+                subtitles[lang].append({
+                    'url': closed_caption['url'],
+                    'ext': mimetype2ext(closed_caption['content_type']),
+                })
 
         return {
             'id': video_id,
@@ -227,8 +376,32 @@ class YahooIE(InfoExtractor):
             'description': clean_html(meta['description']),
             'thumbnail': meta['thumbnail'] if meta.get('thumbnail') else self._og_search_thumbnail(webpage),
             'duration': int_or_none(meta.get('duration')),
+            'subtitles': subtitles,
         }
 
+    def _get_info(self, video_id, display_id, webpage):
+        region = self._search_regex(
+            r'\\?"region\\?"\s*:\s*\\?"([^"]+?)\\?"',
+            webpage, 'region', fatal=False, default='US').upper()
+        formats = []
+        info = {}
+        for fmt in ('webm', 'mp4'):
+            query_result = self._download_json(
+                'https://video.media.yql.yahoo.com/v1/video/sapi/streams/' + video_id,
+                display_id, 'Downloading %s video info' % fmt, query={
+                    'protocol': 'http',
+                    'region': region,
+                    'format': fmt,
+                })
+            info = self._extract_info(display_id, query_result, webpage)
+            formats.extend(info['formats'])
+        formats.extend(self._extract_m3u8_formats(
+            'http://video.media.yql.yahoo.com/v1/hls/%s?region=%s' % (video_id, region),
+            video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False))
+        self._sort_formats(formats)
+        info['formats'] = formats
+        return info
+
 
 class YahooSearchIE(SearchInfoExtractor):
     IE_DESC = 'Yahoo screen search'